CN113743645A - Online education course recommendation method based on path factor fusion - Google Patents
Online education course recommendation method based on path factor fusion Download PDFInfo
- Publication number
- CN113743645A CN113743645A CN202110810736.5A CN202110810736A CN113743645A CN 113743645 A CN113743645 A CN 113743645A CN 202110810736 A CN202110810736 A CN 202110810736A CN 113743645 A CN113743645 A CN 113743645A
- Authority
- CN
- China
- Prior art keywords
- course
- path
- courses
- weight
- adjacent
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000004927 fusion Effects 0.000 title claims abstract description 7
- 238000004364 calculation method Methods 0.000 claims abstract description 15
- 238000010606 normalization Methods 0.000 claims description 5
- 230000008092 positive effect Effects 0.000 abstract description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
- G06Q50/2053—Education institution selection, admissions, or financial aid
Abstract
The invention provides an online education course recommendation method based on path factor fusion, which fuses path weight factors of adjacent courses, improves a calculation value of user similarity, and introduces a self-adaptive weight threshold into a recommendation method, thereby improving the recommendation accuracy of an online education course recommendation system and recommending courses in line with professional learning directions to users. Finally, a set of recommendation system which takes the user target and interest as guidance and accords with the learning path is formed, and the recommendation system has very positive effect and profound and remote significance for improving the learning efficiency of students.
Description
Technical Field
The invention relates to the field of intelligent recommendation, in particular to a method for recommending online education courses based on path factor fusion.
Background
With the continuous development of information technology, online education uses advanced technical means to move a classroom from offline to online, the requirements and the limitations on time and regions in the traditional education mode are overcome, and meanwhile, the opportunity that a user can acquire education resources more conveniently is realized. However, with the increasing multiplication of internet teaching resources, users face new problems and confusion in online course selection. Firstly, learning has the characteristic of path, and related courses must be learned according to a knowledge path, otherwise, the learning effect is seriously influenced by wrong course sequence; secondly, fragmented learning can lead to fragmented curriculum systems, which are not conducive to users building correct knowledge structures.
Recommendation algorithms in the prior art are mature, but are often applied to application environments with non-path factors, such as commodity recommendation and the like. The advantages of such recommendation algorithms are reflected in: the method can efficiently recommend the content preferred by the user to the user by methods such as interest model matching, content analysis and the like, thereby saving a great deal of time cost. However, the conventional recommendation algorithm cannot produce a very precise effect when recommending courses for a user, because learning among courses has a certain path direction, and the conventional recommendation algorithm does not have the characteristic of sensing a learning path, so that courses which should be learned next in a current learning state are not considered by a system, and courses associated with learning before and after are difficult to recommend. Therefore, higher requirements are put forward on course recommendation modes in online education, and recommended courses are required to conform to certain learning paths.
The invention aims to solve the problems of blindness in learning of students, negligence in course resources and path direction required by the students in course learning by combining online education and recommendation technologies, and forms a learning mode of taking a target as a guide for a user so as to exert the characteristics of a course recommendation system.
Disclosure of Invention
In order to solve the technical problem, the invention provides an intelligent course recommendation method based on path factor fusion in an online education application environment. The recommendation method comprises the following steps:
step 1) counting a learned course set I of the user on a main learning path;
step 2) obtaining a candidate recommended non-school course set J closest to the school courses;
step 3) calculating the adjacent course path weight Route between all the learned course sets I and the candidate unvoiced course sets Ji,j(ii) a Recording the path weight of the directly adjacent course i and course j as Routei,jCounting the total number Num of users who successively learn the two adjacent coursesi,jRoute weight of adjacent coursei,jIs Numi,jThe calculation formula is as follows:
Routei,j＝Numi,j
step 4) normalizing the weight of each path, keeping the value range of the weight between 0 and 1 through normalization, and normalizing
Wherein GetRoutei,jThe normalized weights of the paths of the adjacent courses are obtained, and MaxRoute and MinRoute respectively represent the maximum value and the minimum value of the paths of the adjacent courses;
step 5), calculating the average value of the weight values of each path after normalization treatment:
where Threshold represents the average of the path weights of adjacent courses, GetRoutei,jAnd representing the normalized path weight of the adjacent courses, and N represents the total number of the courses.
Step 6) counting directed paths starting from the learned course set I and directly connecting to the candidate recommended unvoiced course set J;
step 7), calculating to obtain a compensation value M (i, j) of the path weight of the adjacent courses:
when the path weight of the adjacent course is greater than the threshold, the compensation weight is GetRoutei,j+1, and when it is less than the threshold, the compensation weight is
If the two courses are not directly adjacent, the compensation weight is 1;
step 8) calculating to obtain the similarity between courses, adding the compensation weight calculated in the step 7) into the course similarity calculation to increase the influence of the path weight and reduce the influence of the hot courses on the course similarity, wherein the similarity between the courses is calculated by the following formula:
wherein sim (i, j) represents the similarity of course i and course j, and M (i, j) represents the compensation weight;
and 9) counting the sum of the similarity of the unvoiced courses j and all the courses in the learned course set I to obtain the preference degree of the user to the recommended unvoiced courses j, and recommending the courses to the user according to the similarity sequence.
Further, the backbone learning path is constructed by the following method:
step 1.1) integrating the whole learning paths of the historical data set users, constructing the whole learning path of each user according to the time sequence of the learning courses of each user in the historical data set, and integrating the whole learning paths of all the users;
step 1.2) carrying out denoising treatment on the integrated learning whole path; in the denoising process, only the adjacent course paths with the weight greater than or equal to a set threshold value are reserved;
and step 1.3) integrating the learning full path subjected to denoising processing to obtain a main learning path.
The online education course recommendation method provided by the invention integrates the weight factors of adjacent course paths, improves the calculation value of the similarity of users, and introduces the self-adaptive weight threshold value into the recommendation method, thereby improving the recommendation accuracy of the online education course recommendation system and recommending courses which accord with the professional learning direction to the users. Finally, a set of recommendation system which takes the user target and interest as guidance and accords with the learning path is formed, and the recommendation system has very positive effect and profound and remote significance for improving the learning efficiency of students.
Drawings
FIG. 1: the invention discloses a flow chart of a recommendation method for online education courses
The specific implementation mode is as follows:
the present invention will be described in further detail with reference to the following embodiments. FIG. 1 shows a flow of an online education course recommendation method of the present invention. In the course recommendation method of the present invention, first, a backbone learning path is constructed.
For convenience of description, the related terms in the present invention are described as follows:
adjacent course path: is defined as learning the connection between two courses which are successively and adjacently arranged in front of and behind each other in a set time window.
Adjacent course path weights: it is defined as the number of people on the path of two consecutive and adjacent lessons learned within a set time window.
Course learning full path for a user: and obtaining the course learning path of the user according to the time sequence relation among the user behavior data.
A historical user set: all user data sets in the data set that learn about a professional related course.
A main learning path: to describe the backbone learning path between the professional related lessons.
The method for constructing the backbone learning path comprises the following steps:
step 1.1, integrating the whole learning path of the historical data set user. According to the time sequence of the learning courses of each user in the historical user set, the whole course learning path of each user is constructed, and then the whole course learning paths of all the users are integrated.
And 1.2, learning denoising treatment in the whole path. Due to the fact that some users have obvious and unreasonable situations of mistakenly selecting courses, paths which are obviously lower than the path weight in the normal range exist in the whole learning path, for example, the path weight is 1 or 2. Therefore, in the invention, the class paths with lower weight of the adjacent class paths are regarded as noise data, and noise reduction processing is required. In the denoising process, only the adjacent course paths with the weight greater than or equal to a certain set threshold value are reserved. Therefore, reliability of recommendation based on historical user learning directions is improved, and a main learning path is further highlighted.
Step 1.3, branch path processing. In the branch path processing, only the weight value within a certain proportion range of the maximum adjacent course path weight value is kept, so that the path within the maximum weight value fluctuation range can be highlighted.
And obtaining a main learning path through the three steps.
Next, the intelligent course recommendation method based on path factor fusion of the present invention specifically includes the following steps:
(1) and counting a learned course set I of a user on a main learning path according to the current learning state of the user, wherein a specific learned course is marked as I.
(2) And obtaining a candidate recommended non-school course set J closest to the learned courses, wherein the specific candidate recommended non-school courses are marked as J. Specifically, the calculation may be performed according to algorithms such as a collaborative filtering algorithm based on the project, which is not limited in the present invention.
(3) Count the path weight of adjacent courses (called the 'weight' for short)
Recording the path weight of the directly adjacent course i and course j as Routei,jCounting the total number Num of users who successively learn the two adjacent coursesi,jRoute weight of adjacent coursei,jIs Numi,jThe calculation formula is as follows:
Routei,j＝Numi,j(formula-1)
(4) Normalizing adjacent course path weights
Obtain the path weight Route of the adjacent coursesi,jAnd then, normalizing each adjacent course path weight value. The value range of the weight is kept at 0,1 by normalization]And obtaining a normalized formula according to the formula. Wherein GetRoutei,jAnd the normalized weights of the paths of the adjacent courses represent the maximum value and the minimum value in the paths of the adjacent courses respectively by MaxRoute and MinRoute.
(5) Adaptive neighboring course path weight threshold
In the existing collaborative filtering algorithm based on items, the formula for calculating the similarity does not consider the influence of the path on course recommendation, so that most of recommended courses in a recommendation list of a recommendation system are courses which are popular at present and preferred by users in a concentrated manner. The more popular courses are recommended more easily, which results in high similarity among most popular courses, but cannot really explain the high similarity among courses. Therefore, before the similarity is calculated, the path compensation value is designed according to the path weight threshold of the adjacent courses, and the courses are divided into two types. When the course path weight is larger than the threshold, performing relevant compensation on the similarity of the course path weight and the threshold, and improving the contribution of the adjacent course path weight to the similarity; when the adjacent course path weight is less than the adjacent course path weight threshold, the compensation for the correlation degree is further reduced.
It is noted that the selection of the neighboring course path weight threshold directly determines Routei,jThe degree of contribution of (c). In order to avoid the artificial misjudgment of the threshold value, the invention adopts a self-adaptive method to set the threshold value. Namely, the normalized path weights of all adjacent courses are averaged, rather than being simply set manually. The threshold calculation formula is as follows.
Where Threshold represents the adjacent course path weight Threshold, GetRoutei,jAnd representing the normalized path weight of the adjacent courses, and N represents the total number of the courses.
(4) Starting from the learned course set I, directly connecting to the directed paths in the candidate recommended unvoiced course set J, and setting the compensation value of the path weight of the adjacent courses
Because the path weights of different adjacent courses are different, the compensation value is added into the similarity calculation formula, and the influence component of the path factors is improved. When the path weight of the adjacent course is greater than the threshold, the compensation weight is GetRoutei,j+1, and when it is less than the threshold, the compensation weight is
If two courses are not directly adjacent, their compensation weight is 1. In summary, the greater the weight of the adjacent course path, i.e., the greater the number of people learning to select the adjacent course path, the higher the contribution to the similarity calculation of the two courses, and the greater the likelihood that the subsequent course will be recommended to the user.
(5) Adding the compensation weight to the course similarity calculation
The invention adds the compensation weight to the similarity calculation to increase the influence of the path weight and reduce the influence of hot courses on the course similarity, and the improved similarity calculation formula is shown as follows, wherein sim (i, j) represents the similarity between the course i and the course j, and M (i, j) represents the compensation weight.
(6) And counting the sum of the similarity of the unvoiced courses j and all the courses in the learned course set I to obtain the preference degree of the user to the recommended unvoiced courses j, and recommending the courses to the user according to the similarity sequence.
The online education course recommendation method provided by the invention integrates the weight factors of adjacent course paths, improves the calculation value of the similarity of users, and introduces the self-adaptive weight threshold value into the recommendation method, thereby improving the recommendation accuracy of the online education course recommendation system and recommending courses which accord with the professional learning direction to the users. Finally, a set of recommendation system which takes the user target and interest as guidance and accords with the learning path is formed, and the recommendation system has very positive effect and profound and remote significance for improving the learning efficiency of students.
The foregoing is a more detailed description of the present invention that is presented in conjunction with specific embodiments, and the practice of the invention is not to be considered limited to those descriptions. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all should be considered as belonging to the protection scope of the invention.
Claims (2)
1. An online education course recommendation method based on path factor fusion is characterized by comprising the following steps:
step 1) counting a learned course set I of the user on a main learning path;
step 2) obtaining a candidate recommended non-school course set J closest to the school courses;
step 3) calculating the adjacent course path weight Route between all the learned course sets I and the candidate unvoiced course sets Ji，j(ii) a Recording the path weight of the directly adjacent course i and course j as Routei，jCounting the total number Num of users who successively learn the two adjacent coursesi，jRoute weight of adjacent coursei，jIs Numi，jThe calculation formula is as follows:
Routei，j＝Numi，j
step 4) normalizing the weight of each path, keeping the value range of the weight between 0 and 1 through normalization, and normalizing
Wherein GetRoutei，jThe normalized weights of the paths of the adjacent courses are obtained, and MaxRoute and MinRoute respectively represent the maximum value and the minimum value of the paths of the adjacent courses;
step 5), calculating the average value of the weight values of each path after normalization treatment:
where Threshold represents the average of the path weights of adjacent courses, GetRoutei，jAnd representing the normalized path weight of the adjacent courses, and N represents the total number of the courses.
Step 6) counting directed paths starting from the learned course set I and directly connecting to the candidate recommended unvoiced course set J;
step 7), calculating to obtain a compensation value M (i, j) of the path weight of the adjacent courses:
when the path weight of the adjacent course is greater than the threshold, the compensation weight is GetRoutei，j+1, and when it is less than the threshold, the compensation weight is
If the two courses are not directly adjacent, the compensation weight is 1;
step 8) calculating to obtain the similarity between courses, adding the compensation weight calculated in the step 7) into the course similarity calculation to increase the influence of the path weight and reduce the influence of the hot courses on the course similarity, wherein the similarity between the courses is calculated by the following formula:
wherein sim (i, j) represents the similarity of course i and course j, and M (i, j) represents the compensation weight;
and 9) counting the sum of the similarity of the unvoiced courses j and all the courses in the learned course set I to obtain the preference degree of the user to the recommended unvoiced courses j, and recommending the courses to the user according to the similarity sequence.
2. The method for recommending an online education course according to claim 1, wherein: the main learning path is constructed by the following method:
step 1.1) integrating the whole learning paths of the historical data set users, constructing the whole learning path of each user according to the time sequence of the learning courses of each user in the historical data set, and integrating the whole learning paths of all the users;
step 1.2) carrying out denoising treatment on the integrated learning whole path; in the denoising process, only the adjacent course paths with the weight greater than or equal to a set threshold value are reserved;
and step 1.3) integrating the learning full path subjected to denoising processing to obtain a main learning path.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110810736.5A CN113743645B (en) | 2021-07-16 | 2021-07-16 | Online education course recommendation method based on path factor fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110810736.5A CN113743645B (en) | 2021-07-16 | 2021-07-16 | Online education course recommendation method based on path factor fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113743645A true CN113743645A (en) | 2021-12-03 |
| CN113743645B CN113743645B (en) | 2024-02-02 |
Family
ID=78728739
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110810736.5A Active CN113743645B (en) | 2021-07-16 | 2021-07-16 | Online education course recommendation method based on path factor fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113743645B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103544663A (en) * | 2013-06-28 | 2014-01-29 | Tcl集团股份有限公司 | Method and system for recommending network public classes and mobile terminal |
| CN105389622A (en) * | 2015-10-20 | 2016-03-09 | 西安交通大学 | Multi-constraint learning path recommendation method based on knowledge map |
| US20160275636A1 (en) * | 2014-03-17 | 2016-09-22 | Michael Olenick | Method and system for curriculum management services |
| CN106023015A (en) * | 2016-05-18 | 2016-10-12 | 腾讯科技（深圳）有限公司 | Course learning path recommending method and device |
| CN109447869A (en) * | 2018-12-06 | 2019-03-08 | 安徽教育网络出版有限公司 | The recommended method of course in a kind of online education |
| CN109816494A (en) * | 2019-01-31 | 2019-05-28 | 北京卡路里信息技术有限公司 | A kind of course recommended method, device, equipment and storage medium |
| CN109902128A (en) * | 2019-01-17 | 2019-06-18 | 平安科技（深圳）有限公司 | Learning path planing method, device, equipment and storage medium based on big data |
| KR102030149B1 (en) * | 2018-11-20 | 2019-10-08 | 주식회사 화성 | Method for recommending a customized curriculum and software education system by using the method |
| CN110990691A (en) * | 2019-11-14 | 2020-04-10 | 泰康保险集团股份有限公司 | Online course recommendation method and device and computer storage medium |
| US20200234606A1 (en) * | 2019-01-22 | 2020-07-23 | International Business Machines Corporation | Personalized educational planning based on user learning profile |
| CN111581529A (en) * | 2020-05-07 | 2020-08-25 | 之江实验室 | Course recommendation method and device combining student fitness and course collocation degree |
| CN112163119A (en) * | 2020-09-29 | 2021-01-01 | 姜锡忠 | Big data online education platform course optimization method and system |
| CN112749805A (en) * | 2021-01-15 | 2021-05-04 | 浙江工业大学 | Online course recommendation method based on multiple entity relationships |
| US11024190B1 (en) * | 2019-06-04 | 2021-06-01 | Freedom Trail Realty School, Inc. | Online classes and learning compliance systems and methods |
| CN113065342A (en) * | 2021-03-22 | 2021-07-02 | 浙江工业大学 | Course recommendation method based on incidence relation analysis |
-
2021
- 2021-07-16 CN CN202110810736.5A patent/CN113743645B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103544663A (en) * | 2013-06-28 | 2014-01-29 | Tcl集团股份有限公司 | Method and system for recommending network public classes and mobile terminal |
| US20160275636A1 (en) * | 2014-03-17 | 2016-09-22 | Michael Olenick | Method and system for curriculum management services |
| CN105389622A (en) * | 2015-10-20 | 2016-03-09 | 西安交通大学 | Multi-constraint learning path recommendation method based on knowledge map |
| CN106023015A (en) * | 2016-05-18 | 2016-10-12 | 腾讯科技（深圳）有限公司 | Course learning path recommending method and device |
| KR102030149B1 (en) * | 2018-11-20 | 2019-10-08 | 주식회사 화성 | Method for recommending a customized curriculum and software education system by using the method |
| CN109447869A (en) * | 2018-12-06 | 2019-03-08 | 安徽教育网络出版有限公司 | The recommended method of course in a kind of online education |
| CN109902128A (en) * | 2019-01-17 | 2019-06-18 | 平安科技（深圳）有限公司 | Learning path planing method, device, equipment and storage medium based on big data |
| US20200234606A1 (en) * | 2019-01-22 | 2020-07-23 | International Business Machines Corporation | Personalized educational planning based on user learning profile |
| CN109816494A (en) * | 2019-01-31 | 2019-05-28 | 北京卡路里信息技术有限公司 | A kind of course recommended method, device, equipment and storage medium |
| US11024190B1 (en) * | 2019-06-04 | 2021-06-01 | Freedom Trail Realty School, Inc. | Online classes and learning compliance systems and methods |
| CN110990691A (en) * | 2019-11-14 | 2020-04-10 | 泰康保险集团股份有限公司 | Online course recommendation method and device and computer storage medium |
| CN111581529A (en) * | 2020-05-07 | 2020-08-25 | 之江实验室 | Course recommendation method and device combining student fitness and course collocation degree |
| CN112163119A (en) * | 2020-09-29 | 2021-01-01 | 姜锡忠 | Big data online education platform course optimization method and system |
| CN112749805A (en) * | 2021-01-15 | 2021-05-04 | 浙江工业大学 | Online course recommendation method based on multiple entity relationships |
| CN113065342A (en) * | 2021-03-22 | 2021-07-02 | 浙江工业大学 | Course recommendation method based on incidence relation analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113743645B (en) | 2024-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tian et al. | A consensus process based on regret theory with probabilistic linguistic term sets and its application in venture capital | |
| US20210334665A1 (en) | Text-based event detection method and apparatus, computer device, and storage medium | |
| CN110321361B (en) | Test question recommendation and judgment method based on improved LSTM neural network model | |
| CN103943107B (en) | A kind of audio frequency and video keyword recognition method based on Decision-level fusion | |
| WO2019165678A1 (en) | Keyword extraction method for mooc | |
| CN112650886B (en) | Cross-modal video time retrieval method based on cross-modal dynamic convolution network | |
| CN109598995A (en) | Intelligent tutoring system based on Bayes's knowledge trace model | |
| CN111444391B (en) | Video learning achievement evaluation method based on artificial intelligence | |
| CN112733533A (en) | Multi-mode named entity recognition method based on BERT model and text-image relation propagation | |
| CN112990464B (en) | Knowledge tracking method and system | |
| CN110941720A (en) | Knowledge base-based specific personnel information error correction method | |
| CN113361791A (en) | Student score prediction method based on graph convolution | |
| CN114254093A (en) | Multi-space knowledge enhanced knowledge graph question-answering method and system | |
| CN111144462B (en) | Unknown individual identification method and device for radar signals | |
| CN110309515B (en) | Entity identification method and device | |
| CN115761408A (en) | Knowledge distillation-based federal domain adaptation method and system | |
| WO2020135054A1 (en) | Method, device and apparatus for video recommendation and storage medium | |
| CN115270807A (en) | Method, device and equipment for judging emotional tendency of network user and storage medium | |
| CN114885216A (en) | Exercise pushing method and system, electronic equipment and storage medium | |
| CN113743645A (en) | Online education course recommendation method based on path factor fusion | |
| CN117390409A (en) | Method for detecting reliability of answer generated by large-scale language model | |
| CN112559759A (en) | Method and equipment for identifying error relation in knowledge graph | |
| CN117350330A (en) | Semi-supervised entity alignment method based on hybrid teaching | |
| CN115687657A (en) | Knowledge graph-assisted test question recommendation method | |
| CN114912549B (en) | Training method of risk transaction identification model, and risk transaction identification method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |